← All models

Gemini 4 Hydrogen

Language Available Full comparison ↗

google-deepmind/gemini-4-hydrogen · by Google DeepMind · decoder-only-transformer

Pricing — 1 offering(s)

Google DeepMind

About this offering: Gemini 4 Hydrogen, Google's fast Gemini 4 model for everyday coding and agentic work, generally available on the Gemini API since 2026-11-12, on both the Free and Paid tiers. One price at every prompt length up to its 1M-token context, with no >200K long-context band. Google presents it as the upgrade from Gemini 3.8 Flash, which moves to the same $1.50/$7.50 price on 2027-01-01.

Input tokens

  • $1.50 / 1M tokens (input) Current 2026-11-12 → present

Cached input tokens

  • $0.075 / 1M tokens (cached input) Current 2026-11-12 → present

    Implicit caching only (explicit caching is not supported); 95% off input.

Output tokens

  • $7.50 / 1M tokens (output) Current 2026-11-12 → present

    Includes thinking tokens.

Batch

Batch input tokens

  • $0.75 / 1M tokens (input) Current 2026-11-12 → present

Batch output tokens

  • $3.75 / 1M tokens (output) Current 2026-11-12 → present

Showing the active price and any recorded history. Full pricing history is available via the paid API — see API docs.

See how Gemini 4 Hydrogen fits into a cost-aware routing setup

See how →

Capability profile

How Gemini 4 Hydrogen rates across core capability dimensions, with the task-level evidence behind each rating.

reasoning Strong

Four tunable thinking levels (low, medium, high by default, and a new xhigh) for deeper reasoning when a task needs it, per Google's What's new page.

coding Strong

Google's release notes: built for everyday coding and agentic workflows. Google says it writes cleaner code and cuts fabricated claims by more than 5x against Gemini 3.8 Flash (vendor-reported). No independent coding benchmark recorded yet.

tool use Strong

Function calling, code execution, computer use, file search and Google Search/Maps grounding are listed as supported on Google's model page. Google says it batches read and search tool calls in parallel and adapts better when tool calls fail.

instruction following Strong
context window Strong

1,048,576-token unified input and output window; 65,536 output tokens by default, up to about 1M with long decode.

multilingual Strong
speed Strong

Google positions it as the fast Gemini 4 model, for parallel worker and verifier agents alongside Argon and Carbon at lower latency.

cost efficiency Strong

$1.50/$7.50 per 1M tokens at every prompt length, below Gemini 4 Argon ($2/$10 introductory, $4/$20 from 2026-12-01). Gemini 3.8 Flash is cheaper until 2026-12-31 ($0.75/$3.75) and the same price from 2027-01-01.

Ratings are estimated — limited independent data is available for this model.

Operator guidance

Google's recommended upgrade from Gemini 3.8 Flash for coding and agentic work. Through 2026-12-31, 3.8 Flash is half the price ($0.75/$3.75), so test Hydrogen's quality gain against that; from 2027-01-01 both cost $1.50/$7.50 and Hydrogen is the natural default. Step up to Gemini 4 Argon, or Carbon for the hardest tasks, when Hydrogen falls short.

Use cases

Limitations

Citations